Volume 41, Issue 1, Pages 27-33 (January 2004) BACE1 Deficiency Rescues Memory Deficits and Cholinergic Dysfunction in a Mouse Model of Alzheimer's Disease  Masuo Ohno, Evgeny A Sametsky, Linda H Younkin, Holly Oakley, Steven G Younkin, Martin Citron, Robert Vassar, John F Disterhoft  Neuron  Volume 41, Issue 1, Pages 27-33 (January 2004) DOI: 10.1016/S0896-6273(03)00810-9 Copyright © 2004 Cell Press Terms and Conditions

Figure 1 BACE1 Null Mutation Rescues Memory Deficits in the Tg2576 Alzheimer's Model (A) Social recognition memory assessed with a 3 hr intertrial delay (n = 10–20). The amount of investigation time during the second exposure to the same juvenile mouse divided by that of the initial investigation time × 100 (% investigation) was used as an index of social recognition memory. Note that only the Tg2576+ group does not show a reduction in spontaneous investigation to a familiar juvenile (approximately 100%) and thus is significantly impaired in this hippocampus-dependent test. (B) Spontaneous alternation Y maze performance for the measurement of spatial working memory (n = 12–21). Tg2576+ mice perform poorly (only slightly above 50% chance levels) in the Y maze test as compared to wild-type control and BACE1−/−·Tg2576+ mice. BACE1−/− mice are moderately but significantly impaired. (C) Total number of arm entries reflecting exploratory activity in the Y maze (n = 12–21). Note that both BACE1−/− and BACE1−/−·Tg2576+ mice explore more than wild-type mice. Each column represents the mean ± SEM. Significant differences from wild-type group (*p < 0.05, **p < 0.01) and Tg2576+ group (#p < 0.05, ##p < 0.01), compared by ANOVA and post hoc Fisher's PLSD test. Neuron 2004 41, 27-33DOI: (10.1016/S0896-6273(03)00810-9) Copyright © 2004 Cell Press Terms and Conditions

Figure 2 BACE1 Null Mutation Rescues Hippocampal Cholinergic Dysfunction in the Tg2576 Alzheimer's Model (A) AHP in response to a 100 ms depolarizing current injection sufficient to elicit a burst of 7 action potentials was recorded from hippocampal CA1 pyramidal cells. Representative traces show the post-burst AHP before (control) and after the application of 0.5 μM carbachol (CCh). CCh at 0.5 μM selectively inhibits the slow component of AHP (sAHP) without affecting the peak amplitude of AHP. Note that the effect of CCh on sAHP in Tg2576+ neurons is reduced as compared to hippocampal neurons from the other three groups. (B and C) Summary bar graphs showing CCh-induced reduction in sAHP measured by amplitudes at 1 s (B) and by integrated areas between 0.3 s and 4.8 s (C) after pulse offset. For both measures, the reduction of Tg2576+ sAHP values following CCh application is less than that of wild-type and BACE1−/−·Tg2576+. Each column represents the mean ± SEM of post-CCh sAHP expressed as % of control (pre-CCh) levels (n = 5–10). Significant differences from wild-type group (*p < 0.05) and Tg2576+ group (#p < 0.05), compared by ANOVA and post hoc Fisher's PLSD test. Neuron 2004 41, 27-33DOI: (10.1016/S0896-6273(03)00810-9) Copyright © 2004 Cell Press Terms and Conditions

Figure 3 Relationship between Cerebral Aβ Levels and Cognitive Performance Social recognition is impaired by excessive levels of Aβ in Tg2576+ mice but is not affected by ablation of Aβ in BACE1−/− mice. In contrast, spontaneous alternation performance in the Y maze is sensitive to disruption by both excess (Tg2576+) and deficient (BACE1−/−) levels of cerebral Aβ. Importantly, BACE1−/−·Tg2576+ bigenic mice are unimpaired in both hippocampal learning tasks, most likely because brain Aβ levels have been restored to normal. Neuron 2004 41, 27-33DOI: (10.1016/S0896-6273(03)00810-9) Copyright © 2004 Cell Press Terms and Conditions